Tidyverse as standard

[PhDyellow]

Meeting with Ant today, he called me out for suggesting reshape2::dcast() instead of the Tidyverse spread().

@jaseeverett, is the default requirement that the lab use Tidyverse for everything unless the individual has a good reason?

A good reason would be a 10GB data frame that can only be handled by data.table or offloading the data.frame to disk.

[jaseeverett]

I think we should advocate the use of tidyverse for small/medium programming jobs - tidying up data, etc. I do believe it is a useful environment, everything is compatible, it is much more readable and there are books out there that focus on it. Using other random packages can be harder to learn and get help.

BUT having said that, I completely agree with you that it is not a perfect solution and there will be plenty of occasions where other solutions are better.

For a new lab member I would probably say, we think the tidy verse is the best way to start, but please use whatever tools are most efficient for your project.

[jaseeverett]

So in answer to @PhDyellow question I would say it is a defacto standard, and recommended for us, but not the only option. Feel free to explore other options.

[jaseeverett]

I wish Ant would hurry up and register for GitHub so we can include him in the discussions

[PhDyellow]

Yes, I agree, tell people to start with tidyverse, then switch to other tools when it is clear tidyverse is not appropriate. For consistency within the lab, I think people should at least put in a comment in their code or github issue when they move away, to make it clear why they stopped using tidyverse for a task.